This surprised me. So many threads here have dealt with the disproportonate number of African American ballots spoiled in Florida. As it turns out there was a correlation between race and spoilage - but the most victimized were African American Republicans. Would anyone still claiming that the Republicans “stole the election” like to explain these numbers?
Could it be true that, as the authors conclude, “those who screamed discrimination the loudest may have the most to hide?” The numbers seem to suggest that. “In excess of 50 times” constitutes a political “badge of fraud,” IMO.
The “badges of fraud” are circumstances which, in the absence of direct evidence, suggest a fraud was perpetrated - usually used in the context of financial transactions. A search on “badges of fraud” would probably turn up several thousand hits.
The Lott-Glassman study can be understood in relation to the report of the US Civil Rights Commission, which found that disproportionately many Black votes had been discarded in Florida. That study used a type of statistical inference, which depends on a number of assumptions – assumptions that cannot be verified. One support the USCRC offered for their conclusion was that race explained the distribution of discarded votes better than any other variable they looked at.
Now there is more data. Thanks to the individual review of all discarded ballots, for most of them it is now known whether they were intended for Gore or for Bush. Using this additional data, Lott and Glassman found a much better explanatory variable, i.e., Black Republican voter, rather than just Black voter.
One reason the L-G makes intuitive sense is that IIRC most Blacks live in Democratic-controlled districts. Why should Democrats want to throw out Black votes? But, it makes a lot more sense for Democrats to throw out Black Republican votes.
Please produce evidence that “Democrats” were able to distinguish “Black” votes from all others, and also evidence that “Democrats” intentionally and disproportionately “threw out” “Black” “Republican” votes while “throwing out” proportionately fewer “Republican” votes of other races.
The study will stand or fall on its own. I haven’t seen it. If anyone has a link to it, please share it with us.
In the absence of the study, we can formulate hypotheses and debate their plausibility. The idea that Republicans unfairly invalidated many Black Democratic votes would be one such hypothesis. I offered an alternative. MG, you know I didnt ssay I had evidence. I said that my hypothesis makes more sense than the other hypothesis. YMMV
Here’s a more detailed hypothesis that could conceivably account for the results of the Lott-Glassman study. Suppose that:[li]Most Black voters live in Democratic districts. []Blacks were more likely than whites to submit a flawed ballot []Some Democratic officials were harsher in throwing out Bush votes than Gore votes. That is, Bush votes with a slight flaws were thrown out; Gore votes with slight flaws were counted.[/list] Then, it might follow that a disproportionately high percentage of Black Republicans could have had their votes not counted.[/li]
As you and I both said, I don’t have evidence for this hypothesis. In fact, parts of it seem far-fetched. (However, if the L-H study is valid, then something strange must have occurred.)
I hope the L-H study becomes available soon, while this thread is still live. Then we can debate the study on its merits.
Isn’t it sad that you can compare “African American Republicans” to “average African Americans,” of whom “African American Republicans” are presumably a part, and discover such a disparity?
Easily explained, Gadarene. Blacks were far more likely to live in precincts with crappy voting equipment. But I’m sure december will be along shortly to tell you how it can be interpreted to show that Democrats are racists.
minty, I guess you’re questioning the internal consistency of my two statements. I said my hypothesis makes more sense than the other one. It seems reasonable IMHO that a higher % of Black Rep votes got thrown out than Black Dem votes. However, a 50 to one ratio seems hard to swallow.
Another hypothesis is that there could have been one particular district with major hanky-panky, so that the votes thrown out there dominated the statistics. Or, the number of Black Republicns may be so small, that the ratio isn’t as meaningful. As my statistician wife points out, if there were just two Black Republican and one had his vote thrown out, that would be 50%. If 1% of Black Dems had their votes thrown out, then you’d have a 50 to one ratio.
All these different ideas just reinforce my desire to see the actual study. We might disagree technically with the study. At the very least the study would show which conceivable explanations might actually be possible.
Since the statistical issues have come up, also note that the very small number of Floridian African-American Republicans as compared to African-American Democrats (about 1 R for every 20 D’s) means that the numbers have disproportionate effects for the two groups. As an example, consider a pool of 101 voters consisting of 100 women and 1 man, where everybody has an equal small chance of spoiling a ballot, and say that one of the women spoiled a ballot. Headline: “1% of Female Ballots Discarded”. Now imagine that one woman and the lone man both spoiled ballots. Headline: “100% of Male Ballots Discarded! Chances of Disenfranchisement for Men One Hundred Times Greater than for Women!!” Etc. etc. etc. Discrepancies in the sample sizes can magnify discrepancies in the results. [Note added in preview: this is what december said in his last post.]
The LA Times seems to have applied some additional statistical processing to the existing data without really thinking about its logical implications, as minty pointed out. I don’t think it’s really possible to draw conclusions from these results without having more information about them.
december:In the absence of the study, we can formulate hypotheses and debate their plausibility.
Yeah, who needs the facts anyway? They just limit our creative freedom of interpretation. Seriously, I think the term “sterile debate” was coined for situations just like this. We simply don’t have enough information about the data to know what it really means.
This, by the way, is what I’m suspecting actually happened.
(Thanks for clearing that up, minty! You can’t imagine the dastardliness I inferred from december’s comment. Why, I was about ready to demonize him! :eek: )
That “little nugget” comes from Mary Frances Berry, head of the U.S. Commission on Civil Rights. It was the major result of the earlier study that she reported on. I was more moderate than she was. I stated it as a hypothesis; she stated it as a conclusion. See: http://www.usccr.gov/vote2000/stdraft1/main.htm
minty wrote:
Minty, do you have a cite?
Incidentally, the USCCR study specifically explored the relationship of voting equipment to discarded ballots. The study said that the relation between race and flawed ballots was stronger than the relation between voting equipment and flawed ballots. In other words, their study said that Black votes were disproportionately thrown out, and that voting equipment does not account for the difference.
This is probably true. A very high percentage of the Black voters were first time voters - I believe an aggressive get-out-the-vote effort almost doubled black voting (as a percentage of the total vote) in FL as compared to the previous presidential election (or maybe it was the previous election).
Where is this implied? The issue is whether Democrats are able to distinguish Republican votes from all others.
(Having said all that, I am skeptical of the study myself)
Hey minty! I’m interested in that little nugget again!
“Why december,” he said wonderingly, “Whatever do you think could account for this correlation? Why would blacks spoil more ballots than whites, if not for faulty equipment?”
Hmmm?
Interestingly, I read the statistical study on which the report was based. I remember clearly what it said. This may be a case where the published summary and/or the press release differed from the underlying study.
It’s too late for me to keep at this, but perhaps I’ll look for the passage over the next couple of days.
Seriously, I think the term “sterile debate” was coined for situations just like this. We simply don’t have enough information about the data to know what it really means.